JOURNAL OF CLINICAL MICROBIOLOGY, June 1990, p. 1417-14210095-1137/90/061417-05$02.00/0

Detection and Characterization of Fecal Verotoxin-ProducingEscherichia coli from Healthy Cattle

MARIA A. MONTENEGRO,l* MICHAEL BULTE,2 THORSTEN TRUMPF,2 STOJANKA ALEKSIC,3GERHARD REUTER,2 EBERHARD BULLING,1 AND REINER HELMUTH'

Institute for Veterinary Medicine (Robert von Ostertag Institute), Federal Health Office,' and Institute forMeat Hygiene, Free University of Berlin,2 D-1000 Berlin 33, and Institute for Hygiene,

Health Service Hamburg, D-2000 Hamburg 28,3 Federal Republic of Germany

Received 30 November 1989/Accepted 26 March 1990

Verotoxin-producing Escherichia coli isolates from feces of healthy cattle were identified by DNA hybrid-ization with verotoxin 1- and verotoxin 2-specific gene probes. Among 259 animals investigated, 28 (10.8%)were found to carry verotoxin-producing E. coli strains. Characterization of the verotoxin-producing isolatesrevealed a heterogeneous population in terms of serotype and toxin type. Nearly 40% of the strains belongedto serogroups known to be pathogenic for humans, i.e., 022, 039, 082, 091, 0113, 0116, 0126, and 0136.Two isolates from different bulls were identified as serotype 0157:H7. Results obtained in this study indicatethat cattle may be an important source of verotoxigenic E. coli involved in human disease.

Verotoxin-producing Escherichia coli (VTEC) strains ofdifferent serotypes are increasingly isolated from cases ofhuman and animal diseases (1, 2, 7, 10, 11, 33). Most of thembelong to the serotype 0157:H7, which can cause hemor-rhagic colitis (HC), hemolytic uremic syndrome (HUS), andthrombotic thrombocytopenic purpura in humans (11). Sev-eral outbreaks have occurred in the last 6 years. In 1983,Riley et al. (22) reported an outbreak in the United Stateswhich was due to the consumption of hamburgers, involving26 patients with HC. In Canada, another infection, also dueto the consumption of hamburgers, led to 73 cases of HC,with 17 fatalities (5). Raw milk was the source of anotheroutbreak in Canada, where 48 persons were treated foreither HC or HUS (4). Consequently, cattle are regarded asa possible reservoir of these highly virulent strains (4, 14, 18;R. Clark, S. McEwen, N. Harnett, H. Lior, and C. Gyles,Abstr., Annu. Meet. Am. Soc., Microbiol. 1988, P48, p.282). In 1988, Morgan et al. (16) reported the first communityoutbreak of HC in the United Kingdom; 24 persons wereaffected, 1 of whom died. In this outbreak, vegetables werethought to have been the source of infection. In a recentpublication, Bockemuhl et al. (2) reported the first food-borne infection due to VTEC in the Federal Republic ofGermany.As a contribution to the epidemiology of VTEC in Ger-

many, we focused our work on cattle as a possible reservoirof these strains. This was achieved by analyzing E. coliisolates originating from fecal samples of animals which areused for milk and meat production in West Berlin. Fecal E.coli isolates from bulls and dairy cows were collected andanalyzed by DNA hybridization with specific gene probesfor verotoxin 1 (VT1) and VT2. Positive strains were sero-typed and confirmed for toxin production by the Vero cellassay.

MATERIALS AND METHODS

Isolation of E. coli strains. Fecal samples of 212 bulls and47 dairy cows were investigated (Table 1). Male animals

* Corresponding author.

came from different regions of the German DemocraticRepublic. The female animals were derived from differentsites in the Federal Republic of Germany, but they werehoused together in West Berlin for various periods (1 to 7months) before sample collection.The fecal samples were homogenized (10 g of feces in 90

ml of NaCi-peptone water) not more than 2 h after havingbeen taken aseptically from the colons of the slaughteredbulls and from the rectums of the dairy cows. Serial dilutions(0.1 ml) were plated on plate count Monensin KCl medium(19), supplemented with 50 ,ug of 4-methyl-umbelliferyl-P-D-glucuronid (Sigma Chemical Co.) per ml or HC mediumas described by Szabo et al. (30). Presumptive E. colicolonies were biochemically confirmed by Gram staining,the cytochrome oxidase test, and a modified IMViC proce-dure (20). The resulting 1,387 confirmed E. coli isolates werestored in microdilution plates at -70°C in L broth (DifcoLaboratories) containing 20% glycerol (15).

Serotyping. Serological analyses were performed by slideand tube agglutination with adsorbed O and H antiseraagainst 170 O and 56 H antigens by following the scheme ofOrskov and Orskov (17).VT assay. Isolates were tested for VT production by the

method of Konowalchuk et al. (12). Strains were grownovernight at 37°C in L broth. Supernatants were obtained bycentrifuging the cultures at 10,000 x g for 15 min and wereevaluated for toxic activity immediately.DNA probes. The VT1-specific gene probe was a 750-

base-pair HincII fragment obtained from the recombinantplasmid NTP705 (32), and the VT2 probe was an 850-base-pair AvaI-PstI fragment from the recombinant plasmidNTP707 (31). These fragments were purified from agarosegels by electroelution. The gene probes were labeled by nicktranslation either with 32P or with biotin by using nicktranslation kits from Amersham Corp. or GIBCO BethesdaResearch Laboratories, respectively.DNA hybridization experiments. E. coli strains were

screened for the presence of VT genes by colony hybridiza-tion (8). Isolates were grown in L broth in microdilutionplates at 37°C overnight and transferred to lactose agarplates with a multipoint inoculator. After overnight incuba-

1417

Vol. 28, No. 6

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

cm o

n 13

Feb

ruar

y 20

22 b

y 58

.152

.226

.171

.

1418 MONTENEGRO ET AL.

TABLE 1. Frequency of VT genes in fecal E. coliisolates from bovines

Group No. (%) of No. of No. (%) of(no. of animals) VT+ animals isolates VT+ isolates

Dairy cows (47) 8 (17.0) 181 21 (11.6)BuIls (212) 20 (9.4) 1,206 36 (3.0)Total (259) 28 (10.8) 1,387 57 (4.1)

tion at 37°C, grown colonies were further transferred toSchwarzband filters (no. 586; Schleicher & Schuell, Inc.).The filters were processed by the method of Maas (13).Hybridizations with 32P-labeled probes were performed as

described previously (15), except that the hybridizationtemperature was 42°C. For hybridizations with biotin-la-beled DNA probes, processed filters were previously treatedwith proteinase K (Bethesda Research Laboratories; 1 mg/ml) by the method of Sethabutr et al. (25). Hybridizationexperiments with biotinylated probes and detection of posi-tive colonies were carried out as recommended by theinstructions of the manufacturer (BluGene; Bethesda Re-search Laboratories).For preparation of chromosomal DNA, bacteria were

lysed by the method of Hull et al. (9). DNA was purified byphenol extraction and ethanol precipitation and dissolved in10 mM Tris hydrochloride, pH 7.6-lmM EDTA. Digestionwith the restriction endonuclease HindIII (Bethesda Re-search Laboratories) was performed as recommended by thesupplier. Southern blots (27) were prepared as describedpreviously (15). Hybridization experiments using biotiny-lated probes were carried out as described above.

RESULTS

Distribution of VT-related sequences in fecal E. coli isolatesfrom healthy cattle. A total of 1,387 E. coli isolates obtainedfrom 259 animals were analyzed by colony hybridizationwith VT1- and VT2-specific gene probes. Table 1 shows thenumbers and frequencies of gene probe-positive isolates,distributed according to the sex of the animals. A total of 21(11.6%) of the 181 isolates from dairy cows and 36 (3.0%) ofthe 1,206 isolates from bulls hybridized with one or bothprobes. In all, 57 strains (4.1%) carrying one or both VTgenes could be identified in 28 animals (10.8%).The distribution of VT genes according to serotype is

shown in Table 2. Twenty-six strains (45.6%) were found tocarry both toxin genes. Hybridization to the VT1 probealone was observed in only nine isolates (15.8%). Homologyto the VT2 gene probe alone was found in 22 isolates(38.6%).

Production of VT by bovine E. coli strains carrying VTgenes. Culture supernatants of the 57 strains that hybridizedwith the VT probes were tested for cytotoxicity in the Verocell assay. All of them elicited the typical cytopathic effectdescribed for VTs. In contrast, none of 24 arbitrarily chosen,probe-negative isolates produced VT. Toxin titers weredetermined as the highest dilution giving a cytotoxic effecton Vero cells after 24 h of incubation. No differences in toxinproduction could be associated with a particular toxin type;toxin titers ranged from 102 to >105. It can be concluded thatall isolates that hybridized to the VT probes were VTECstrains.

Southern blot analysis of VTEC strains. In order to deter-mine the location of VT genes, total DNA was extractedfrom some strains producing VT1 or both toxins, digested

TABLE 2. Distribution of VT1 and VT2 genes in VTEC strainsfrom bovines according to serotype

No. of: Hybridization with

Serotype probe for:Strains Animals VT1 VT2

03:H- 2 1 + -O10:H21a 2 1 +022:H8 2 1 - +039:H4Ob 1 1 - +075:H8 1 1 - +082:H8c 6 4 + +082:H8c 1 1 - +082:H40c 1 1 + +091:H10 1 1 - +0104:H21 1 1 + +0105:H18 1 1 + -0113:H21 6 4 - +0116:H21c 15 10 + +0116:H21 1 1 - +0126:H20 2 1 - +0126:H21 1 1 - +0136:Hl2a 4 2 +0139:H8b 1 1 - +0156:H21b 1 1 + +0157:H7 2 2 - +Orough:H18 2 1 + +O?:H16d 1 1 - +O?:H29 2 1 - +

a Five strains with these characteristics were isolated from one bull.b Strains isolated from one bull.C Five strains with these characteristics were isolated from one cow.d This serovar possesses a new O antigen, which is currently under

investigation.

with the restriction endonuclease HindIII, and separated onan agarose gel. The DNA fragments were blotted ontonitrocellulose and probed with VT1 and VT2 (Fig. 1). VT1hybridized with two HindIII fragments in each case (Fig.1A). Since strains belonging to the same serotype showed acommon hybridization pattern, apart from sharing otherproperties, such as plasmid profiles, biovars (not shown),and toxin genes, we conclude that they were related oridentical isolates. In contrast to that, we could observedifferences in the locations of VT1 genes when strainsbelonging to different serotypes were analyzed. Strainsbelonging to serotypes 0116:H21 and Orough:H18 shared a

common HindIII fragment of 3.2 kilobases, while strains ofserotype 03:H- had totally different hybridizing fragmentsof 10.7 and 6.7 kilobases (Fig. 1A). Hybridization to VT2sequences always occurred to large HindIII fragments ofabout 20 kilobases (Fig. 1B). This was also found in strainsthat only produced VT2 (data not shown).

Serotyping of VTEC isolates. Fifty-four VTEC strainscould be classified into 19 defined serotypes (Table 2). Forthe other three isolates, the O antigen was not typeable.Almost 58% of all the strains belonged to only four sero-

types, namely 0116:H21 (n = 16), 082:H8 (n = 7), 0113:H21 (n = 6), and 0136:H12 (n = 4). The rest of the strainswere distributed among 19 other serotypes. With two excep-tions, strains of serotypes 082:H8 and 0116:H21 producedboth VT1 and VT2. Isolates that belonged to serotype0113:H21 carried only VT2 genes, whereas only VT1 se-

quences were found in all four strains of serotype 0136:H12.VTEC belonging to serotypes 082:H8 and 0113:H21 (n =

12) have been associated with human infections (11). Otherpotentially pathogenic strains of serogroups 022, 039, 091,0126, and 0157 (n = 9) were found in small numbers. All of

J. CLIN. MICROBIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

cm o

n 13

Feb

ruar

y 20

22 b

y 58

.152

.226

.171

.

VT1 AND VT2 IN E. COLI FROM CATTLE 1419

A 1 2 3 4 5 6 7 8

_»O-- - _

B ) 1 2 3 4 5 6 7 8

4.t

FIG. 1. Southern blot analysis of VTEC belonging to threedifferent serotypes. (A) Chromosomal blot probed with a biotiny-lated VT1 probe. (B) The same blot as in panel A probed with a

biotinylated VT2 probe. Lanes 1, NTP707 cut with Aval and PstI(arrow indicates fragment used as the probe); 2 through 7, HindIII-digested chromosomal DNA from E. coli 2478 (Orough:H18) (lane2), E. coli 2477 (Orough:H18) (lane 3), E. coli 2431 (0116:H21) (lane4), E. coli 2430 (0116:H21) (lane 5), E. coli 2405 (03:H-) (lane 6), E.coli 2403 (03:H-) (lane 7); 8, NTP705 cut with HincII (probefragment is indicated by the arrow; other fragments are partialproducts of digestion).

them produced VT2 only. Among the latter strains, twoisolates from two different bulls were identified as belongingto the high-virulence serotype 0157:H7. In all, 22 (38.6%)isolates belonged to E. coli serogroups known to be patho-genic.Some animals were found to carry more than one VTEC

strain. The simultaneously occurring strains were unrelatedto each other. Strains of serotypes 082:H8, 082:H40, and0116:H21 were isolated from the same animal (Table 2).Strains of serotypes 039:H40, 0139:H8, and 0156:H21 were

isolated from another animal. Also, 010:H21 and 0136:H12isolates were found together in a third bovine.

DISCUSSION

The present study has shown that VTEC, the causativeagent of HC, HUS, and thrombotic thrombocytopenic pur-pura, could be identified in fecal E. coli isolates from healthycattle in the Federal Republic of Germany and the German

Democratic Republic. To our knowledge, this is the firstepidemiological survey on the occurrence of these patho-genic strains in healthy cattle in Germany. Workers in othercountries, namely in Canada, the United States, the UnitedKingdom, and Sri Lanka, have postulated cattle to be areservoir of VTEC strains (4, 11, 18, 26; Clark et al., Abstr.Annu. Meet. Am. Soc. Microbiol.). Only a few sporadiccases of HC, HUS, and thrombotic thrombocytopenic pur-pura have been reported in the Federal Republic of Germany(2, 10, 28, 33). One of these cases (2) involved serotype022:H8, which we also found in our investigation.The high incidence of VTEC strains found in cattle poses

the question of whether humans are at a risk of acquiringthese infections. Fecal contamination of meat during theprocess of slaughter cannot be totally ruled out. Therefore,the presence of VTEC in food is likely to occur. Investiga-tions by Doyle and Schoeni (6) have shown a detection rateof VTEC in 3.7% of meat samples of bovine origin. Further-more, a lower rate of contamination (between 1.5 and 2%)has been described by the same authors for food of porcine,avian, or sheep origin. Their data, however, referred only toE. coli serotype 0157:H7. Our recent data on the serologicaldetection of this serotype in more than 1,000 strains derivedfrom over 100 food samples failed to show any positive strain(data not shown). To what extent other VTEC strains arepresent in food samples is the subject of our present work.The rates of VTEC isolation from cattle described here are

very similar to those found by Clark et al. (Abstr. Annu.Meet. Am. Soc. Microbiol.) in Canada, the country wherethe most food-borne infections due to VTEC have beendocumented. They found VTEC strains in 19.5% of theinvestigated dairy cows, the mean value for the total cattlepopulation being 11.6%. These workers detected serotype0157:H7 in only one cow and three bulls from a total of 600animals. Together with our data, these findings support thenotion that cattle are an important reservoir of VTEC.Nearly 40% of our isolates belonged to serogroups or

serotypes recognized as being pathogenic for humans (2, 11).Most ofthem, i.e., 039:H40, 082:H8, 091:H10, 0113:H21,0126:H20, 0126:H21, and 0157:H7, were isolated fromcattle for the first time in this country. The lack of epidemi-ological data on VTEC other than 0157:H7 probably resultsfrom the widespread use of sorbitol-supplemented media, onwhich only strains from serotype 0157:H7 can be presump-tively detected.The risk of infection is also supported by the finding that

all probe-positive isolates were able to express VT, asmeasured in the Vero cell test. Strains expressing either VT1or VT2 are known to be pathogenic for humans (24). None ofthe animals carrying VTEC strains presented any obviousclinical symptoms. Smith et al. (26) were able to detectVTEC strains in cattle and swine with enteric diseases.These strains, however, belong to serogroups other than theones reported here. These authors identified several 026:Hl strains in cattle, a serotype that has been also isolatedfrom a sporadic case of HC (3). More recently, Appel et al.(1) also identified (in diseased pigs) VTEC strains of sero-groups similar to the ones reported by Smith et al. (26). Itseems likely that some VTEC serogroups are associatedwith infections of humans and animals, whereas others arefound in infections of either humans or animals.

In some cases, different animals were found to carryidentical isolates; others, in contrast, were found to harborseveral different VTEC strains. For this reason, we assumethat the detection of potential VTEC strains can only besuccessful if a sufficient number of E. coli isolates from each

VOL. 28, 1990

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

cm o

n 13

Feb

ruar

y 20

22 b

y 58

.152

.226

.171

.

1420 MONTENEGRO ET AL.

fecal sample is examined. A mean of four to six isolates fromeach sample was investigated in this study, which in ouropinion seems to be the lowest acceptable limit. Ideally, 10isolates from each sample should be analyzed in order toincrease the probability of identifying VTEC strains amongthe far more frequently occurring commensal strains.Two chromosomal HindIlI fragments were found to hy-

bridize to the VT1 gene probe in Southern blot analyses ofsix arbitrarily chosen strains from three different serotypes(Fig. 1A). This was to be expected, since the HincII frag-ment used as a probe contains a HindIlI recognition site (31,32). However, when chromosomal blots from strains belong-ing to different serotypes were compared, different pairs ofHindIII fragments hybridized in each case. This might reflectthe presence of three different bacteriophages in the sero-types analyzed. Different phages carrying VT1 genes havealready been demonstrated in serotypes 0157:H7 and 026:Hil (23, 29). More recently, Rietra et al. (21) demonstratedthe existence of diverse phages in VTEC strains fromdifferent serogroups.

ACKNOWLEDGMENTS

We thank Herlinde Irsigler, Daniela Marshel, Edith Burbes, andIngrid Jetschmann for valued technical assistance. We also thank H.Smith for providing NTP705- and NTP707-carrying strains.

This work was supported by research grants of the FederalMinistry for Research and Technology, Bonn, Federal Republic ofGermany (project no. 0318953 A) to G. Reuter.

LITERATURE CITED

1. Appel, G., C. Ewald, A. Heer, G. v. Mickwitz, S. Aleksic, H.Russmann, T. Meyer, and H. Karch. 1989. Vorkommen undBedeutung von Verotoxin-(Shiga-like-Toxin-) produzierendenEscherichia-Coli-Stammen beim Schwein. Tierarztl. Umsch.44:410-420.

2. Bockemuhl, J., H. Karch, H. Russmann, S. Aleksic, R. Wiss, andP. Emmrich. 1990. Shiga-like toxin (verotoxin)-produzierendeEscherichia coli 022:H8. Bundesgesundheitsblatt. 33:3-6.

3. Bopp, C. A., K. D. Greene, F. P. Downes, E. G. Sowers, J. G.Wells, and I. K. Wachsmuth. 1987. Unusual verotoxin-pro-ducing Escherichia coli associated with hemorrhagic colitis. J.Clin. Microbiol. 25:1486-1489.

4. Borczyk, A. A., M. A. Karmali, H. Lior, and L. M. C. Duncan.1987. Bovine reservoir for verotoxin-producing Escherichia coli0157:H7. Lancet i:98.

5. Carter, A. O., A. A. Borczyk, J. A. K. Carlson, B. Harvey, J. C.Hockin, M. A. Karmali, C. Krishnan, D. A. Korn, and H. Lior.1987. A severe outbreak of Escherichia coli 0157:H7-associatedhemorrhagic colitis in a nursing home. N. Engl. J. Med.317:1496-1500.

6. Doyle, M. P., and J. L. Schoeni. 1987. Isolation of Escherichiacoli 0157:H7 from retail fresh meats and poultry. Apple. Envi-ron. Microbiol. 53:2394-2396.

7. Gonzalez, E. A., and J. Blanco. 1985. Production of cytotoxinVT in enteropathogenic and non-enteropathogenic Escherichiacoli strains of porcine origin. FEMS Microbiol. Lett. 26:127-130.

8. Grunstein, M., and D. S. Hogness. 1975. Colony hybridization: amethod for the isolation of cloned DNAs that contain a specificgene. Proc. Natl. Acad. Sci. USA 72:3961-3965.

9. Hull, R. A., R. E. Gill, P. Hsu, B. H. Minshew, and S. Falkow.1981. Construction and expression of recombinant plasmidsencoding type 1 or D-mannose-resistant pili from a urinary tractinfection Escherichia coli isolate. Infect. Immun. 33:933-938.

10. Karch, H., and J. Bockemuhl. 1989. Infektionen durch entero-hfmorrhagische Escherichia coli (EHEC): ein klinisches undmikrobiologisches Problem und eine Herausforderung fur denoffentlichen Gesundheitsdienst. Immun. Infekt. 17:206-211.

11. Karmali, M. A. 1989. Infection by Verocytotoxin-producingEscherichia coli. Clin. Microbiol. Rev. 2:15-38.

12. Konowalchuk, J., J. I. Speirs, and S. Stavric. 1977. Veroresponse to a cytotoxin of Escherichia coli. Infect. Immun.18:775-779.

13. Maas, R. 1983. An improved colony hybridization method withsignificantly increased sensitivity for detection of single copygenes. Plasmid 10:296-298.

14. Martin, M. L., L. D. Shipman, M. E. Potter, I. K. Wachsmuth,J. G. Wells, K. Hedberg, R. V. Tauxe, J. P. Davis, J. Arnoldi,and J. Tilleli. 1986. Isolation of Escherichia coli 0157:H7 fromdairy cattle associated with two cases of haemolytic uraemicsyndrome. Lancet ii:1043.

15. Montenegro, M. A., G. J. Boulnois, and K. N. Timmis. 1984.Molecular epidemiology by colony hybridization using clonedgenes, p. 92-103. In A. Pühler and K. N. Timmis (ed.),Advanced molecular genetics. Springer-Verlag KG, Berlin.

16. Morgan, G. M., C. Newman, S. R. Palmer, J. B. Allen, W.Shepherd, A. M. Rampling, R. E. Warren, R. J. Gross, S. M.Scotland, and H. R. Smith. 1988. First recognized communityoutbreak of haemorrhagic colitis due to verotoxin-producingEscherichia coli 0157.H7 in the UK. Epidemiol. Infect. 101:83-91.

17. 0rskov, F., and I. 0rskov. 1984. Serotyping of Escherichia coli,p. 43-112. In T. Bergan (ed.), Methods in microbiology, vol. 15.Academic Press, Inc. (London), Ltd., London.

18. 0rskov, F., I. 0rskov, and J. A. Villar. 1987. Cattle as reservoirof verotoxin-producing Escherichia coli 0157:H7. Lancet ii:276.

19. Petzel, J. P., and P. A. Hartman. 1985. Monensin-based mediumfor determination of total gram-negative bacteria and Esche-richia coli. Appl. Environ. Microbiol. 49:925-933.

20. Powers, E. M., and T. G. Latt. 1977. Simplified 48-hour IMVICtest: an agar plate method. Appl. Environ. Microbiol. 41:274-279.

21. Rietra, P. J. G. M., G. A. Willshaw, H. R. Smith, A. M. Field,S. M. Scotland, and B. Rowe. 1989. Comparison of Vero-cytotoxin-encoding phages from Escherichia coli of human andbovine origin. J. Gen. Microbiol. 135:2307-2318.

22. Riley, L. W., R. S. Remis, S. D. Helgerson, H. B. Mc Gee, J. G.Wells, B. R. Davis, R. J. Herbert, E. S. Olcott, L. M. Johnson,N. T. Hargrett, P. A. Blake, and M. L. Cohen. 1983. Hemor-rhagic colitis associated with a rare Escherichia coli serotype.N. Engl. J. Med. 308:681-685.

23. Scotland, S. M., H. R. Smith, G. A. Willshaw, and B. Rowe.1983. Vero cytotoxin production in strain of Escherichia coli isdetermined by genes carried on bacteriophage. Lancet ii:216.

24. Scotland, S. M., G. A. Willshaw, H. R. Smith, and B. Rowe.1987. Properties of strains of Escherichia coli belonging toserogroup 0157 with special reference to production of Verocytotoxins VT1 and VT2. Epidemiol. Infect. 99:613-624.

25. Sethabutr, O., S. Hanchalay, P. Echeverria, D. N. Taylor, and U.Leksomboon. 1985. A non-radioactive DNA probe to identifyShigella and enteroinvasive Escherichia coli in stools of chil-dren with diarrhoea. Lancet ii:1095-1097.

26. Smith, H. R., S. M. Scotland, G. A. Willshaw, C. Wray, I. M.McLaren, T. Cheasty, and B. Rowe. 1988. Vero cytotoxinproduction and presence of VT genes in Escherichia coli strainsof animal origin. J. Gen. Microbiol. 134:829-834.

27. Southern, E. M. 1975. Detection of specific sequences amongDNA fragments separated by gel electrophoresis. J. Mol. Biol.89:503-517.

28. Stenger, K. O., F. Windler, H. Karch, H. v. Wulffen, and J.Heesemann. 1988. Hemolytic uremic syndrome associated withinfection by verotoxin producing Escherichia coli 0111 ina woman on oral contraceptives. J. Clin. Nephrol. 29:153-158.

29. Stockbine, N., L. R. M. Marques, J. W. Newland, H. W. Smith,R. K. Homes, and A. D. O'Brien. 1986. Two toxin-convertingphages from Escherichia coli 0157:H7 strain 933 encode anti-genically distinct toxins with similar biologic activities. Infect.Immun. 53:135-140.

30. Szabo, R. A., E. C. D. Todd, and A. Jean. 1986. Method to

J. CLIN. MICROBIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

cm o

n 13

Feb

ruar

y 20

22 b

y 58

.152

.226

.171

.

VT1 AND VT2 IN E. COLI FROM CATTLE 1421

isolate Escherichia coli 0157:H7 from food. J. Food Protect.49:768-772.

31. Willshaw, G. A., H. R. Smith, S. M. Scotland, A. M. Field, andB. Rowe. 1987. Heterogeneity of Escherichia coli phages encod-ing Vero cytotoxins: comparison of cloned sequences determin-ing VT1 and VT2 and development of specific gene probes. J.

Gen. Microbiol. 133:1309-1317.32. Willshaw, G. A., H. R. Smith, S. M. Scotland, and B. Rowe.

1985. Cloning of genes determining the production of Vero-cytotoxin by Escherichia coli. J. Gen. Microbiol. 131:3047-3053.

33. Wuthe, H.-H. 1987. Escherichia coli 0157:H7 bei hamorrhagis-cher Colitis. Dtsch. Med. Wochenschr. 112:1761-1762.

VOL. 28, 1990

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

cm o

n 13

Feb

ruar

y 20

22 b

y 58

.152

.226

.171

.